Background: Imatinib, a first-generation tyrosine kinase inhibitor (1GTKI), has been the standard of care for patients with chronic myeloid leukemia (CML) for over 10 years. Second-generation TKIs (2GTKI), nilotinib and dasatinib, were approved by the FDA as first-line therapy for CML in 2010. Guidelines recommend first-line therapy with any of these agents, but no consensus exists for determining which agent should be used first. Objective: This study examined the association between initiation of 1GTKI vs. 2GTKI and the following outcomes: medication adherence, health services utilization, and direct health care costs. Method: This was a retrospective cohort study of commercial and Medicare patients identified from the Humana Research Database. Patients newly initiating 1GTKI or 2GTKI therapy between June 1, 2010 and December 31, 2011 who were continuously enrolled for 4 months during the baseline period, between the ages of 18 and 89 years old at the index date, and had a diagnosis for CML were included. Multivariate logistic regression was used to investigate the association between TKI therapy and adherence, defined as proportion of days covered =0.85. Multivariate logistic regression and generalized linear models (GLMs) were used to examine the association between TKI therapies and health services utilization. GLMs were used to examine the association between TKI therapies and direct health care costs (plan and patient paid) during the 12 months follow-up period. Results: 368 patients met the inclusion criteria (1GTKI, n=237; 2GTKI, n=131). There was no difference in adherence between patients initiating a 1GTKI compared to a 2GTKI (odds ratio=0.88, 95% confidence interval [CI] 0.55-1.40). Outpatient visits were the most frequently used health service by both treatment groups; initiating a 2GTKI was associated with increased outpatient visits (incidence rate ratio=1.12, 95% CI 1.06-1.20). There were no statistically significant differences in emergency room visits or inpatient visits between patients initiating a 1GTKI versus 2GTKI but of those patients who had an inpatient visit, initiating 2GTKI therapy was associated with increased inpatient days compared to initiating a 1GTKI (IRR=3.25, 95% CI 2.43-4.35). Total costs were 1.3 times higher for 2GTKIs initiators versus 1GTKI initiators ($86,509 vs. $66,443; P=0.001). Differences in total costs were attributed to significant differences in TKI pharmacy costs. Conclusions: There were no differences in adherence among patients newly initiating 1GTKI vs. 2GTKI therapy. Patients initiating a 2GTKI had increased outpatient visits compared to patients initiating 1GTKI but there was no difference in medical costs between TKI generations. Total and TKI pharmacy costs, were significantly higher among 2GTKI initiators compared to 1GTKI initiators. This cost differential is likely to increase substantially with the impending release of generic imatinib, and payers may begin to use preferred agent strategies. Oncologists will weigh these factors, along with effectiveness and safety data, when choosing a first-line TKI therapy.
